OpenAI partners with Broadcom, TSMC for new in-house chip
- OpenAI is working with Broadcom and TSMC to enhance its chip supply chain amid rising processing demands.
- The company is assembling a team of around 20 chip engineers, aiming to design its first custom chip by 2026.
- This initiative highlights OpenAI's strategy to manage increasing infrastructure costs and secure chip supply, similar to challenges faced by other tech giants.
In a bid to meet the soaring computational demands of artificial intelligence, OpenAI is joining forces with Broadcom and Taiwan Semiconductor Manufacturing Company (TSMC). This collaboration is part of OpenAI's effort to diversify its supply chain and stabilize its infrastructure costs, which have become increasingly critical due to the rapid advancements in AI technology. OpenAI has shifted its focus from building foundries to designing its in-house chips after recognizing the high costs and lengthy timelines associated with chip manufacturing. OpenAI has begun assembling a specialized team of about 20 chip engineers, including veterans who have previously worked on Google's Tensor Processing Units (TPUs). This team is tasked with the ambitious goal of developing OpenAI's first custom chip by 2026; however, the timeline is flexible and subject to change based on the development process's demands. The recent news regarding this partnership has had a positive impact on the stock market, with shares of Broadcom rising by 4% and TSMC's US-traded shares increasing by over 1%. Such partnerships are reflective of a larger trend in the tech industry, where giants like Amazon, Meta, Google, and Microsoft face similar challenges in sourcing chips. With the integration of AMD chips and Nvidia’s GPUs, which are in limited supply, OpenAI's strategic shift represents a critical move to ensure a consistent chip supply. This initiative is vital not just for OpenAI but underscores the ongoing struggles that many tech companies encounter in their AI ventures.